Presentation is loading. Please wait.

Presentation is loading. Please wait.

Automatic Image Alignment via Motion Estimation

Similar presentations


Presentation on theme: "Automatic Image Alignment via Motion Estimation"— Presentation transcript:

1 Automatic Image Alignment via Motion Estimation
15-463: Rendering and Image Processing Alexei Efros …with a lot of slides stolen from Steve Seitz

2 Today Direct (pixel-based) alignment Feature-based alignment Reading:
Brute Force Search Gradient Search (Motion Estimation) Lucas-Kanade Feature-based alignment Interest Points SIFT Brown & Lowe, “Recognising Panoramas” Reading: Szeliski, Sections 3 and 4

3 Image Alignment How do we align two images automatically?
Two broad approaches: Feature-based alignment Find a few matching features in both images compute alignment Direct (pixel-based) alignment Search for alignment where most pixels agree

4 Direct Alignment The simplest approach is a brute force search (hw1)
Need to define image matching function SSD, Normalized Correlation, edge matching, etc. Search over all parameters within a reasonable range: e.g. for translation: for tx=x0:step:x1, for ty=y0:step:y1, compare image1(x,y) to image2(x+tx,y+ty) end; Need to pick correct x0,x1 and step What happens if step is too large?

5 Direct Alignment (brute force)
What if we want to search for more complicated transformation, e.g. homography? for a=a0:astep:a1, for b=b0:bstep:b1, for c=c0:cstep:c1, for d=d0:dstep:d1, for e=e0:estep:e1, for f=f0:fstep:f1, for g=g0:gstep:g1, for h=h0:hstep:h1, compare image1 to H(image2) end; end; end; end; end; end; end; end;

6 Problems with brute force
Not realistic Search in O(N8) is problematic Not clear how to set starting/stopping value and step What can we do? Use pyramid search to limit starting/stopping/step values For special cases (rotational panoramas), can reduce search slightly to O(N4): H = K1R1R2-1K (4 DOF: f and rotation) Alternative: gradient decent on the error function i.e. how do I tweak my current estimate to make the SSD error go down? Can do sub-pixel accuracy BIG assumption? Images are already almost aligned (<2 pixels difference!) Can improve with pyramid Same tool as in motion estimation

7 Motion estimation: Optical flow
Will start by estimating motion of each pixel separately Then will consider motion of entire image

8 Why estimate motion? Lots of uses Track object behavior
Correct for camera jitter (stabilization) Align images (mosaics) 3D shape reconstruction Special effects

9 Problem definition: optical flow
How to estimate pixel motion from image H to image I? Solve pixel correspondence problem given a pixel in H, look for nearby pixels of the same color in I Key assumptions color constancy: a point in H looks the same in I For grayscale images, this is brightness constancy small motion: points do not move very far This is called the optical flow problem

10 Optical flow constraints (grayscale images)
Let’s look at these constraints more closely brightness constancy: Q: what’s the equation? A: H(x,y) = I(x+u, y+v) small motion: (u and v are less than 1 pixel) suppose we take the Taylor series expansion of I:

11 Optical flow equation Combining these two equations
In the limit as u and v go to zero, this becomes exact

12 Optical flow equation Q: how many unknowns and equations per pixel?
Intuitively, what does this constraint mean? The component of the flow in the gradient direction is determined The component of the flow parallel to an edge is unknown A: u and v are unknown, 1 equation This explains the Barber Pole illusion

13 Aperture problem

14 Aperture problem

15 Solving the aperture problem
How to get more equations for a pixel? Basic idea: impose additional constraints most common is to assume that the flow field is smooth locally one method: pretend the pixel’s neighbors have the same (u,v) If we use a 5x5 window, that gives us 25 equations per pixel!

16 RGB version How to get more equations for a pixel?
Basic idea: impose additional constraints most common is to assume that the flow field is smooth locally one method: pretend the pixel’s neighbors have the same (u,v) If we use a 5x5 window, that gives us 25*3 equations per pixel!

17 Lukas-Kanade flow Prob: we have more equations than unknowns
Solution: solve least squares problem minimum least squares solution given by solution (in d) of: The summations are over all pixels in the K x K window This technique was first proposed by Lukas & Kanade (1981)

18 Conditions for solvability
Optimal (u, v) satisfies Lucas-Kanade equation When is This Solvable? ATA should be invertible ATA should not be too small due to noise eigenvalues l1 and l2 of ATA should not be too small ATA should be well-conditioned l1/ l2 should not be too large (l1 = larger eigenvalue) ATA is solvable when there is no aperture problem

19 Local Patch Analysis

20 Edge large gradients, all the same large l1, small l2

21 Low texture region gradients have small magnitude small l1, small l2

22 High textured region gradients are different, large magnitudes
large l1, large l2

23 Observation This is a two image problem BUT
Can measure sensitivity by just looking at one of the images! This tells us which pixels are easy to track, which are hard very useful later on when we do feature tracking...

24 Errors in Lukas-Kanade
What are the potential causes of errors in this procedure? Suppose ATA is easily invertible Suppose there is not much noise in the image When our assumptions are violated Brightness constancy is not satisfied The motion is not small A point does not move like its neighbors window size is too large what is the ideal window size?

25 Iterative Refinement Iterative Lukas-Kanade Algorithm
Estimate velocity at each pixel by solving Lucas-Kanade equations Warp H towards I using the estimated flow field - use image warping techniques Repeat until convergence

26 Revisiting the small motion assumption
Is this motion small enough? Probably not—it’s much larger than one pixel (2nd order terms dominate) How might we solve this problem?

27 Reduce the resolution!

28 Coarse-to-fine optical flow estimation
Gaussian pyramid of image H Gaussian pyramid of image I image I image H u=10 pixels u=5 pixels u=2.5 pixels u=1.25 pixels image H image I

29 Coarse-to-fine optical flow estimation
Gaussian pyramid of image H Gaussian pyramid of image I image I image H run iterative L-K warp & upsample run iterative L-K . image J image I

30 Beyond Translation So far, our patch can only translate in (u,v)
What about other motion models? rotation, affine, perspective Same thing but need to add an appropriate Jacobian (see Table 2 in Szeliski handout):

31 Image alignment Goal: estimate single (u,v) translation for entire image Easier subcase: solvable by pyramid-based Lukas-Kanade

32 Lucas-Kanade for image alignment
Pros: All pixels get used in matching Can get sub-pixel accuracy (important for good mosaicing!) Relatively fast and simple Cons: Prone to local minima Images need to be already well-aligned  What if, instead, we extract important “features” from the image and just align these?

33 Feature-based alignment
Find a few important features (aka Interest Points) Match them across two images Compute image transformation as per HW#2 Choosing Features Choose only the points (“features”) that are salient, i.e. likely to be there in the other image How to find these features? windows where has two large eigenvalues Called the Harris Corner Detector

34 Feature Detection

35 Feature Matching One possibility: Problems:
Match features found in image1 with features found in image2 e.g. SSD of image patches around each feature Use successful matches to estimate homography Do something to get rid of outliers Problems: What if the image patches for several interest points look similar? Make patch size bigger What if the image patches for the same feature look different due to scale, rotation, etc. Use Lucas-Kanade with affine motion model Better solution: Scale-Invariant Feature Transform (SIFT)

36 Invariant Features Schmid & Mohr 1997, Lowe 1999, Baumberg 2000, Tuytelaars & Van Gool 2000, Mikolajczyk & Schmid 2001, Brown & Lowe 2002, Matas et. al. 2002, Schaffalitzky & Zisserman 2002

37 SIFT Features Invariant Features SIFT features are…
Establish invariant frame Find optimal scale Maxima/minima of scale-space DOG  x, y, s Find optimal orientation Maximum of distribution of local gradients  q Form descriptor vector Histogram of smoothed local gradients 128 dimensions SIFT features are… Geometrically invariant to similarity transforms, some robustness to affine change Photometrically invariant to affine changes in intensity

38 Example: Recognising Panoramas
M. Brown and D. Lowe, University of British Columbia

39 Why “Recognising Panoramas”?

40 Why “Recognising Panoramas”?
1D Rotations (q) Ordering  matching images

41 Why “Recognising Panoramas”?
1D Rotations (q) Ordering  matching images

42 Why “Recognising Panoramas”?
1D Rotations (q) Ordering  matching images

43 Why “Recognising Panoramas”?
1D Rotations (q) Ordering  matching images 2D Rotations (q, f) Ordering  matching images

44 Why “Recognising Panoramas”?
1D Rotations (q) Ordering  matching images 2D Rotations (q, f) Ordering  matching images

45 Why “Recognising Panoramas”?
1D Rotations (q) Ordering  matching images 2D Rotations (q, f) Ordering  matching images

46 Why “Recognising Panoramas”?

47 Overview Feature Matching Image Matching Bundle Adjustment
Multi-band Blending Results Conclusions

48 RANSAC for Homography

49 RANSAC for Homography

50 RANSAC for Homography

51 Probabilistic model for verification

52 Finding the panoramas

53 Finding the panoramas

54 Finding the panoramas

55 Finding the panoramas

56 Homography for Rotation
Parameterise each camera by rotation and focal length This gives pairwise homographies

57 Bundle Adjustment New images initialised with rotation, focal length of best matching image

58 Bundle Adjustment New images initialised with rotation, focal length of best matching image

59 Multi-band Blending Burt & Adelson 1983
Blend frequency bands over range  l

60 Results


Download ppt "Automatic Image Alignment via Motion Estimation"

Similar presentations


Ads by Google